A Low Resolution Epistasis Mapping Approach To Identify Chromosome Arm Interactions in Allohexaploid Wheat

Analysis of historical selection in winter wheat

KEY MESSAGE

Modeling of the distribution of allele frequency over year of variety release identifies major loci involved in historical breeding of winter wheat. Winter wheat is a major crop with a rich selection history in the modern era of crop breeding. Genetic gains across economically important traits like yield have been well characterized and are the major force driving its production. Winter wheat is also an excellent model for analyzing historical genetic selection. As a proof of concept, we analyze two major collections of winter wheat varieties that were bred in Western Europe from 1916 to 2010, namely the Triticeae Genome (TG) and WAGTAIL panels, which include 333 and 403 varieties, respectively. We develop and apply a selection mapping approach, Regression of Alleles on Years (RALLY), in these panels, as well as in simulated populations. RALLY maps loci under sustained historical selection by using a simple logistic model to regress allele counts on years of variety release. To control for drift-induced allele frequency change, we develop a hybrid approach of genomic control and delta control. Within the TG panel, we identify 22 significant RALLY quantitative selection loci (QSLs) and estimate the local heritabilities for 12 traits across these QSLs. By correlating predicted marker effects with RALLY regression estimates, we show that alleles whose frequencies have increased over time are heavily biased toward conferring positive yield effect, but negative effects in flowering time, lodging, plant height and grain protein content. Altogether, our results (1) demonstrate the use of RALLY to identify selected genomic regions while controlling for drift, and (2) reveal key patterns in the historical selection in winter wheat and guide its future breeding.

Genomic Prediction Methods Accounting for Nonadditive Genetic Effects

The use of genomic information for prediction of future phenotypes or breeding values for the candidates to selection has become a standard over the last decade. However, most procedures for genomic prediction only consider the additive (or substitution) effects associated with polymorphic markers. Nevertheless, the implementation of models that consider nonadditive genetic variation may be interesting because they (1) may increase the ability of prediction, (2) can be used to define mate allocation procedures in plant and animal breeding schemes, and (3) can be used to benefit from nonadditive genetic variation in crossbreeding or purebred breeding schemes. This study reviews the available methods for incorporating nonadditive effects into genomic prediction procedures and their potential applications in predicting future phenotypic performance, mate allocation, and crossbred and purebred selection. Finally, a brief outline of some future research lines is also proposed.

The decreased expression of GW2 homologous genes contributed to the increased grain width and thousand‑grain weight in wheat-Dasypyrum villosum 6VS·6DL translocation lines

This study demonstrated that the aberrant transcription of DvGW2 contributed to the increased grain width and thousand-grain weight in wheat-Dasypyrum villosum T6VS·6DL translocation lines. Due to the high immunity to powdery mildew, Dasypyrum villosum 6VS has been one of the most successful applications of the wild relatives in modern wheat breeding. Along with the desired traits, side-effects could be brought when large alien chromosome fragments are introduced into wheat, but little is known about effects of 6VS on agronomic traits. Here, we found that T6VS·6DL translocation had significantly positive effects on grain weight, plant heightand spike length, and small negative effects on total spikelet number and spikelet compactness using recipient and wheat-D. villosum T6VS·6DL allohexaploid wheats, Wan7107 and Pm97033. Further analysis showed that the 6VS segment might exert direct genetic effect on grain width, then driving the increase of thousand-grain weight. Furthermore, comparative transcriptome analysis identified 2549 and 1282 differentially expressed genes (DEGs) and 2220 and 1496 specifically expressed genes (SEGs) at 6 days after pollination (DAP) grains and 15 DAP endosperms, respectively. Enrichment analysis indicated that the process of cell proliferation category was over-represented in the DEGs. Notably, two homologous genes, TaGW2-D1 and DvGW2, were identified as putative candidate genes associated with grain weight and yield. The expression analysis showed that DvGW2 had an aberrant expression in Pm97033, resulting in significantly lower total expression level of GW2 than Wan7107, which drives the increase of grain weight and width in Pm97033. Collectively, our data indicated that the compromised expression of DvGW2 is critical for increased grain width and weight in T6VS·6DL translocation lines.

Genetic basis of maize kernel oil-related traits revealed by high-density SNP markers in a recombinant inbred line population

BACKGROUND

Maize (Zea mays ssp. mays) is the most abundantly cultivated and highly valued food commodity in the world. Oil from maize kernels is highly nutritious and important for the diet and health of humans, and it can be used as a source of bioenergy. A better understanding of genetic basis for maize kernel oil can help improve the oil content and quality when applied in breeding.

RESULTS

In this study, a KUI3/SC55 recombinant inbred line (RIL) population, consisting of 180 individuals was constructed from a cross between inbred lines KUI3 and SC55. We phenotyped 19 oil-related traits and subsequently dissected the genetic architecture of oil-related traits in maize kernels based on a high-density genetic map. In total, 62 quantitative trait loci (QTLs), with 2 to 5 QTLs per trait, were detected in the KUI3/SC55 RIL population. Each QTL accounted for 6.7% (qSTOL1) to 31.02% (qBELI6) of phenotypic variation and the total phenotypic variation explained (PVE) of all detected QTLs for each trait ranged from 12.5% (OIL) to 52.5% (C16:0/C16:1). Of all these identified QTLs, only 5 were major QTLs located in three genomic regions on chromosome 6 and 9. In addition, two pairs of epistatic QTLs with additive effects were detected and they explained 3.3 and 2.4% of the phenotypic variation, respectively. Colocalization with a previous GWAS on oil-related traits, identified 19 genes. Of these genes, two important candidate genes, GRMZM2G101515 and GRMZM2G022558, were further verified to be associated with C20:0/C22:0 and C18:0/C20:0, respectively, according to a gene-based association analysis. The first gene encodes a kinase-related protein with unknown function, while the second gene encodes fatty acid elongase 2 (fae2) and directly participates in the biosynthesis of very long chain fatty acids in Arabidopsis.

CONCLUSIONS

Our results provide insights on the genetic basis of oil-related traits and a theoretical basis for improving maize quality by marker-assisted selection.

Genomic mechanisms of climate adaptation in polyploid bioenergy switchgrass

Long-term climate change and periodic environmental extremes threaten food and fuel security¹ and global crop productivity^2-4. Although molecular and adaptive breeding strategies can buffer the effects of climatic stress and improve crop resilience⁵, these approaches require sufficient knowledge of the genes that underlie productivity and adaptation⁶-knowledge that has been limited to a small number of well-studied model systems. Here we present the assembly and annotation of the large and complex genome of the polyploid bioenergy crop switchgrass (Panicum virgatum). Analysis of biomass and survival among 732 resequenced genotypes, which were grown across 10 common gardens that span 1,800 km of latitude, jointly revealed extensive genomic evidence of climate adaptation. Climate-gene-biomass associations were abundant but varied considerably among deeply diverged gene pools. Furthermore, we found that gene flow accelerated climate adaptation during the postglacial colonization of northern habitats through introgression of alleles from a pre-adapted northern gene pool. The polyploid nature of switchgrass also enhanced adaptive potential through the fractionation of gene function, as there was an increased level of heritable genetic diversity on the nondominant subgenome. In addition to investigating patterns of climate adaptation, the genome resources and gene-trait associations developed here provide breeders with the necessary tools to increase switchgrass yield for the sustainable production of bioenergy.

Efficient Algorithms for Calculating Epistatic Genomic Relationship Matrices

The genomic relationship matrix plays a key role in the analysis of genetic diversity, genomic prediction, and genome-wide association studies. The epistatic genomic relationship matrix is a natural generalization of the classic genomic relationship matrix in the sense that it implicitly models the epistatic effects among all markers. Calculating the exact form of the epistatic relationship matrix requires high computational load, and is hence not feasible when the number of markers is large, or when high-degree of epistasis is in consideration. Currently, many studies use the Hadamard product of the classic genomic relationship matrix as an approximation. However, the quality of the approximation is difficult to investigate in the strict mathematical sense. In this study, we derived iterative formulas for the precise form of the epistatic genomic relationship matrix for arbitrary degree of epistasis including both additive and dominance interactions. The key to our theoretical results is the observation of an interesting link between the elements in the genomic relationship matrix and symmetric polynomials, which motivated the application of the corresponding mathematical theory. Based on the iterative formulas, efficient recursive algorithms were implemented. Compared with the approximation by the Hadamard product, our algorithms provided a complete solution to the problem of calculating the exact epistatic genomic relationship matrix. As an application, we showed that our new algorithms easily relieved the computational burden in a previous study on the approximation behavior of two limit models.

Exploring Deep Learning for Complex Trait Genomic Prediction in Polyploid Outcrossing Species

Genomic prediction (GP) is the procedure whereby the genetic merits of untested candidates are predicted using genome wide marker information. Although numerous examples of GP exist in plants and animals, applications to polyploid organisms are still scarce, partly due to limited genome resources and the complexity of this system. Deep learning (DL) techniques comprise a heterogeneous collection of machine learning algorithms that have excelled at many prediction tasks. A potential advantage of DL for GP over standard linear model methods is that DL can potentially take into account all genetic interactions, including dominance and epistasis, which are expected to be of special relevance in most polyploids. In this study, we evaluated the predictive accuracy of linear and DL techniques in two important small fruits or berries: strawberry and blueberry. The two datasets contained a total of 1,358 allopolyploid strawberry (2n=8x=112) and 1,802 autopolyploid blueberry (2n=4x=48) individuals, genotyped for 9,908 and 73,045 single nucleotide polymorphism (SNP) markers, respectively, and phenotyped for five agronomic traits each. DL depends on numerous parameters that influence performance and optimizing hyperparameter values can be a critical step. Here we show that interactions between hyperparameter combinations should be expected and that the number of convolutional filters and regularization in the first layers can have an important effect on model performance. In terms of genomic prediction, we did not find an advantage of DL over linear model methods, except when the epistasis component was important. Linear Bayesian models were better than convolutional neural networks for the full additive architecture, whereas the opposite was observed under strong epistasis. However, by using a parameterization capable of taking into account these non-linear effects, Bayesian linear models can match or exceed the predictive accuracy of DL. A semiautomatic implementation of the DL pipeline is available at https://github.com/lauzingaretti/deepGP/.

Prediction of Subgenome Additive and Interaction Effects in Allohexaploid Wheat

Whole genome duplications have played an important role in the evolution of angiosperms. These events often occur through hybridization between closely related species, resulting in an allopolyploid with multiple subgenomes. With the availability of affordable genotyping and a reference genome to locate markers, breeders of allopolyploids now have the opportunity to manipulate subgenomes independently. This also presents a unique opportunity to investigate epistatic interactions between homeologous orthologs across subgenomes. We present a statistical framework for partitioning genetic variance to the subgenomes of an allopolyploid, predicting breeding values for each subgenome, and determining the importance of inter-genomic epistasis. We demonstrate using an allohexaploid wheat breeding population evaluated in Ithaca, NY and an important wheat dataset from CIMMYT previously shown to demonstrate non-additive genetic variance. Subgenome covariance matrices were constructed and used to calculate subgenome interaction covariance matrices for variance component estimation and genomic prediction. We propose a method to extract population structure from all subgenomes at once before covariances are calculated to reduce collinearity between subgenome estimates. Variance parameter estimation was shown to be reliable for additive subgenome effects, but was less reliable for subgenome interaction components. Predictive ability was equivalent to current genomic prediction methods. Including only inter-genomic interactions resulted in the same increase in accuracy as modeling all pairwise marker interactions. Thus, we provide a new tool for breeders of allopolyploid crops to characterize the genetic architecture of existing populations, determine breeding goals, and develop new strategies for selection of additive effects and fixation of inter-genomic epistasis.

Homeologous Epistasis in Wheat: The Search for an Immortal Hybrid

Hybridization between related species results in the formation of an allopolyploid with multiple subgenomes. These subgenomes will each contain complete, yet evolutionarily divergent, sets of genes. Like a diploid hybrid, allopolyploids will have two versions, or homeoalleles, for every gene. Partial functional redundancy between homeologous genes should result in a deviation from additivity. These epistatic interactions between homeoalleles are analogous to dominance effects, but are fixed across subgenomes through self pollination. An allopolyploid can be viewed as an immortalized hybrid, with the opportunity to select and fix favorable homeoallelic interactions within inbred varieties. We present a subfunctionalization epistasis model to estimate the degree of functional redundancy between homeoallelic loci and a statistical framework to determine their importance within a population. We provide an example using the homeologous dwarfing genes of allohexaploid wheat, Rht-1, and search for genome-wide patterns indicative of homeoallelic subfunctionalization in a breeding population. Using the IWGSC RefSeq v1.0 sequence, 23,796 homeoallelic gene sets were identified and anchored to the nearest DNA marker to form 10,172 homeologous marker sets. Interaction predictors constructed from products of marker scores were used to fit the homeologous main and interaction effects, as well as estimate whole genome genetic values. Some traits displayed a pattern indicative of homeoallelic subfunctionalization, while other traits showed a less clear pattern or were not affected. Using genomic prediction accuracy to evaluate importance of marker interactions, we show that homeologous interactions explain a portion of the nonadditive genetic signal, but are less important than other epistatic interactions.